The cytochrome P450 subfamily CYP3A represents one of the most important families of the P450 superfamily and plays a major role in the metabolism of an ever expanding list of therapeutic compounds (23, 24). This family comprises the most abundantly expressed P450s in human livers, and is responsible for the metabolism of over 50% of all clinically used drugs, including the dihydropyridines, cyclosporin, erythromycin and barbiturates (1). Wide inter-individual variation in the metabolism of CYP3A substrates has been noted and is a factor in determining individual drug efficacy. Evidence also exists for the metabolism of an array of lipophilic environmental pollutants, including the activation of pro-carcinogens such as aflatoxin B1 by members of this subfamily (2).
Presently, four CYP3A cDNAs have been identified in humans, CYP3A3, CYP3A4, CYP3A5 and CYP3A7. It is believed that CYP3A3 represents an allelic variant of CYP3A4, whilst CYP3A4 and CYP3A7 are found only in human adult and fetal livers respectively (3). Initial experiments suggested that a polymorphism existed in CYP3A4 (4). However other studies, whilst confirming a wide range of inter-individual variation in CYP3A4 expression have failed to confirm the original bimodality (5, 6). Overlapping substrate specificities between CYP3A5 and CYP3A5 have previously made it difficult to separate metabolism by these isoforms; consequently little phenotyping data have been produced to study variation in CYP3A5 activity in humans. However, there is evidence for the polymorphic expression of CYP3A5. Use of both immunoblotting and Northern analysis have detected CYP3A5 expression in only 10–30% of human livers (7, 8, 9). More recently, analysis of 30 human liver samples using immunoblotting found that only 3% showed no detectable CYP3A5, whilst a large number had trace amounts, suggesting that a polymorphism in this enzyme may be regulatory as opposed to structural (10). Comparisons of the 5′ flanking regions from the CYP3A4, 3A5 and 3A7 genes have identified putative binding sites for several transcriptional regulatory factors common to all isoforms (11, 12, 13). However, the molecular basis, if any, for this inter-individual variation in expression of the CYP3A sub-family members has so far remained unclear. Indeed it has been suggested that the host cellular environment may be a greater determinant of inducibility than gene structure (14). However, the determination of a major genetic component to variant expression and activity, linked to an easy screening method, would be extremely beneficial, not only in providing a predictor of individual response to drugs which are metabolised by these isoforms, but also in facilitating association studies between CYP3A and disease processes.
The delineation of CYP3A4 and CYP3A5 metabolism has been shown to be possible using the sedative midazolam as a probe drug (15). In this case two metabolites are formed, 1-hydroxy midazolam (1-OHM) and 4-hydroxy midazolam (4-OHM). Those samples containing a higher proportion of CYP3A5 compared to CYP3A4 have their metabolism driven towards the 1-OHM route and therefore show a higher ratio of 1-OHM/4-OHM than those containing only CYP3A4. The present inventors have now established that two polymorphisms, located in putative transcriptional regulatory regions, which caused increased CYP3A5 gene expression and metabolic activity are linked and have developed assays for their detection. These assays will allow prediction of inter-individual variability in response to drugs metabolised by this isoform, as well as facilitating disease association studies.